Genetic/Epigenetic Markers, Social Contexts, Lifecourse &Risky Health Behaviors PROJECT SUMMARY/ABSTRACT This application addresses broad challenge area (01) Behavior, Behavioral Change, and Prevention and two highly related specific Challenge Topics: (A) 01-OD(OBSSR)-102* Methods for studying the interactions among behaviors, environments, and genetic/epigenetic processes and (B) 01-DA-111: Approaches to study the interactions among individual behaviors, social and physical environments, and genetic/epigenetic processes during critical developmental periods. The first specific topic of 01-OD(OBSSR)- 102 emphasizes the development of analytical methods in this area and the second topic of 01-DA-111 emphasizes the actual integration of environmental and developmental variables with genotypic information in order to permit comprehensive model-building and hypothesis testing for determining genetic, environmental, and developmental contributions to substance abuse and related phenotypes. The overall challenge for our application is to integrate genetic polymorphisms, epigenetic markers, social contextual measures, and developmental periods into analysis of risky health behaviors. The past 2-3 years saw an unparalleled succession of discoveries in the genomics of complex diseases. So far, GWAS has focused on estimating genetic main effects. However, the general consensus is that the links between genetic heritage and complex human traits, especially human behaviors, are enormously complicated, typically involving multiple genes, epigenetic markers, social contextual factors, life course stages, and the interactions among these sources. In this application, we propose three goals in response to the overall challenge. First, we propose two novel analytical approaches for addressing the issue of high-dimension-low-sample-size (HDLSS) or too many independent variables for the available sample size, which is one of the most difficult issues in genetic studies of complex human traits. These are the method of variable selection under HDLSS and the method of simultaneous selection of important variables and stratification of subjects. Both methods are for estimating genetic main effects and gene-environment interaction effects. Our second goal is to actually integrate genetic data (an Illumina GoldenGate array of 1536 SNPs), social contextual data, and life course into investigations of the causes of nine most common risky health behaviors (marijuana use, cocaine use, other illegal drug use, number of sexual partners, binge drinking, drinking quantity, smoking quantity, smoking frequency, and seatbelt non-wearing), drawing data (N=2,600) from the National Longitudinal Study of Adolescent Health. Our Third Goal is to carry out a pilot project that investigates the feasibility of saliva DNA for a population-level methylation study. The pilot project addresses several technical issues, which must be resolved before a large- scale population methylation study via saliva DNA can be implemented - the issues of cellular heterogeneity or cell identity, potential contamination of saliva by orally ingested agents, and the detection of informative methylation pattern differences in T cells from Saliva. 1 Genetic/Epigenetic Markers, Social Contexts, Lifecourse &Risky Health Behaviors Public Health Relevance - The overall challenge for our application is to integrate genetic polymorphisms, epigenetic markers, social contextual measures, and developmental periods into analysis of risky health behaviors. This application explores two novel statistical methods designed to address the integration, performs the integrated analysis using Add Health data, and conducts a pilot study investigating the feasibility of using saliva DNA for a large scale epigenetic project.